The Time for Cybernetics Has Come - with Daniel Davies

You know, at the end of this, we have this weird neoliberal non-managerial capitalism where everyone is meant to be managing according to financial markets. Um, or a lot of which was driven by people presuming that financial financial markets were the ultimate information processing system that you could just throw any question at, and you would get an answer coming from the market about anything that you wanted to do.

[Music]

Uh, so I've got with me the economist Dan Davis, who is the author of the uh, unaccountability machine. Uh, we're going to talk about that today. Thanks so much for coming, Dan.

Very much for inviting me. I think this is my first time ever on a live stream.

Is it? Is it actually? Okay. Wow. I'm surprised by that because your book's been doing pretty well, right? I mean, I would have thought you would have got invited to quite a lot of a lot of things.

Um, it's done all right. I've I've been on podcasts, like I've been on so many podcasts. You know, if your computer has started making a funny whining noise, it's probably me on a podcast. Uh, but um, yeah. No, this is the first time with the live technology. I'm very impressed.

Yes. Yeah. Um, it took me about 3 years before I stopped making stupid mistakes on the live streams. Um, I'll say that and I'll I'll have made one. Uh, but yeah, so so this book, I it's so um, it's very much of of the time. I feel like it came along at the right time, and this notion of like accountability, while it's a word we all know and use, right? Um, I think you bring a lot of things together. So I mean, it sort of captures everything from the frustration of, you know, trying to change your broadband provider, which I did recently and like not being able to speak to anyone, right, all the way through to like Trump, right, and like like the 2008 financial crisis. So yeah, could you explain like how you bring those things together with this idea of accountability?

Yeah, I mean, I think the um, I mean the interesting thing to me was that the unaccountability machine was a late change of title. Uh, all the time that I was working on this book, I was calling it Decisions Nobody Made. And okay, it was I mean, I've lately been told that Decisions Nobody's Have Made would be a crap title, and I was very lucky to have changed it. But, um, what what it's all about is what I'd call the industrialization of decision-making. And that's something that's really gone on in the last 50 years, and it's not been well understood or well analyzed at all. Um, so I mean, I kind of like the idea here is that like 100 years ago you might have had a knife and fork which would have been made in Sheffield by a monster cutler, and there would be a human being who would see your cutlery and they'd be able to say, "I made that." And obviously, that's just not how the world works anymore. um, you know, you everything is made by a machine, and you have a machine operator, but that's not the same thing as making the things, but increasingly really since the second world war and at an accelerating rate as organizations have got more and more complicated, as the whole system has got more and more complicated, people have built systems and processes and policies. They've implemented some of them into literal computer software. But in general, most of the people who we still think of as managers and decision-makers, even up to CEO level, even up to quite high in government level, they're actually their daily experience is much more like being the operator of an industrial decision-making machine. You know, they they don't kind of feel like they're actually making these decisions. Because they don't feel like they're making the decisions, it's psychologically intolerable to be held responsible for a decision you didn't make. And so they start creating more and more of these systems and attributing a sort of objective reality to them because that way they can deflect accountability away from themselves into like uh, phrases used in the book is an accountability sink, which is just a social arrangement used in order to direct negative emotion and negative feedback into a place where it can be harmlessly accumulated without causing any trouble to any individual manager. Obviously, the very biggest accountability sink we have is the market, you know, which, you know, I'm just looking at you. I'm about 10 or 20 years older than you, so I I remember when market decisions, or particularly bond market and foreign exchange market decisions, were things that could be criticized. You know, I I remember dimly one of my kind of first adult memories when we got rid of having the the idea that the interest rate was something that was managed by a politically elected representative. Um, these things have all been changed. It has made the world change in a really quite complicated way, and it's all just very very underanalyzed. And you know, in the book, I come up with a couple of theories about why it's been so under analyzed. U most of which basically involve blaming Milton Friedman. That's always that's always a safe bet, isn't it?

I mean, what one thing I one thing I did like about about the book, if you well, one moment I liked in the book, if you could talk us through this, was how you first thought that you'd solved the problem and you thought you were going to be really famous, make loads of money, you know, probably be king of the the universe or something with your new theory and then you realized that you were wrong. So, could you talk us through that cuz I think that personal revelation is quite an interesting way to understand your theory.

I mean, which was actually kind of horrible because like when I had the book, there were decisions nobody made. I had this perfect kind of plan, some it was all going to be the concept of an accountability sink. Have you noticed how everyone blames things on the computer or blames them on the market or blames it on the policy these days? Nobody takes individual accountability for their decisions anymore, and they should, and they're bad people for not doing so. And you know, we should all kind of criticize everyone for that and you know, have a new revolution where more people take accountability for their decisions, and it was a nice clear book. I think it would have sold well in airports. It would have uh, but like one year into writing this thing, I just got more and more questions. I got some very good questions from a very good edible profile books by the way, great publishing house. And I just had this kind of lack of intellectual honesty, which is just, you know, professionally disastrous. It took me the book took another two years to write cuz I was thinking ended up just going, well, actually, firstly, can you really run a modern industrial society based on individual people having individual responsibility for the decisions they make, and if you could, would you want to? You know, a lot of the time the reason that there's a policy in place or that the reason reason that there's a checklist or a system for making this decision is that people who had the power to make it and nominally be accountable for doing so kind of abused that power pretty badly in the past. And so actually, I started, you know, really revising the entire kind of thesis here and thinking, well, what we've got here when we are talking about accountability sinks is ways in which people manage the fact that their environment is just too complicated for them to make sense of. You know, people start basically because the world is too complicated to deal with, people create mental objects and start trying to manage that. Or we kind of put arrangements together to create a shared mental object, and we'll make decisions about our simplified mental object and then we'll hope that those decisions are not too disastrous when they are applied to the real world. And that got me into uh my kind of current intellectual obsession, Stafford Beer and management cybernetics, which is, as far as I can tell, the only time anyone's had a really solid approach at taking a scientific and rigorous attack on the problem of how do you manage overwhelming information flows?

Yeah. Yeah, we we'll talk about that uh uh in a lot of detail. I think one thing that uh one example that popped into my mind when you were talking about that like, you know, the sort of bad decision-makers versus creating a system to make a decision for you was like grades. So everyone complains about grades and you know, it's fair enough, people you know, students are graded too much, everybody knows that it's stressful and all that, but there is evidence, I realized this, that what came before that, which was basically the teacher saying how good you were uh was was worse, it was worse in it was much more biased, discriminatory, who the teacher liked, who looked like the teacher, all that kind of stuff, and now we've got grades and they've got their problems, but they were designed to solve to solve the problem of what came before, right? So, it's just too it would be too easy just to say, you know, Brit, let's go back to the uh what was the example you used right at the beginning with somebody who designed a shoe or something, a blacksmith or something. We can't go back to that kind of decision-making, you you just can't do it, and in many cases, you wouldn't want it.

Yeah. you know, it's sort of which is a lot of the problem kind of skipping ahead a few chapters here with the populist right and the people like the Donald Trump, even Sylvia Biskone, they are to a large extent taking people who are overwhelmed by a world that is basically too complex for them to deal with, partly because you know, government functions which were meant to deal with a lot of that's uncertainty and risk for them are failing to do so. And you got Trump who's just saying, "This is all easy. I can make this simple. I can make, you know, the world manageable and comprehensible for you again." Um, and it's a lie because, you know, to claim you can do that is like saying you can turn back time. It it just can't be done. Unfortunately, we have to live with the world as it is.

I mean, it's interesting because this book, I guess, is about a Did it come out about a year ago or something like that?

um, in earlier this year in the US and in paper book um two years ago, so before the Trump election, which resulted in me saying a few things about it, but my impact on my ability to go to the USA and market it definitely, but it's interesting what he's tried to do now, right? I mean, because I mean, it's difficult when somebody's just obviously a complete but also there's a there's a rationale somewhere there which is that I think he's doing what you were talking about, right? He's trying to turn back time. He's saying, "No, we're going to bring everything under the control of the American people with these tariffs, with whatever political gestures he's making." And clearly, you can't, right? You're seeing that the impact, the outcome is catastrophic.

Yeah. And the problem is everyone around him is trying to absorb all of the chucks from the outside world and deal as a kind of translation and simplification mechanism for him. And you know, I'm talking quite frankly, you know, he's now over the age of 80. Um, his ideological mental world is not as big as it was eight years ago. Yeah. And so he's working on managing this mental representation of America that's in his mind, and it only has very loose connections with the reality of the American economy. And so you've got people, you know, visibly leaping around like uh fleas in a flea circus trying to take the things that he said and convert them into something that can be communicated to the outside world without causing a 10% fall in the US dollar.

That's a really good way of putting it. Yeah. Um, I mean, I was let's let's talk a little bit about technocracy because you do I mean in early in the book you are you do seem to be laying a lot of blame at the feet of technocracy, and you you kind of you you hinted at it then because it's like all so many decisions have been taken out of the hands of the people, and then you get this populist backlash where where it's like a vote, you know, against the elites, against the status quo. And one of the really the things that um stuck with me, I didn't realize this. So both the Maastricht Treaty 3% of GDP deficit number.

Yeah. You're not allowed to deficit over 3% of GDP in the EU.

Yeah. Um, and also the 2% inflation target the central banks across the world have. They're just they're just like almost completely arbitrary. They just came from like no, someone just said them, and now they're like the standard. And it's just like these technocratic decisions uh they seem I mean, is there any justification for this or is it just as ridiculous as it sounds?

Really seem like a good idea for the three the history of the 3% number in the Maastricht stability and growth act is that it was inherited from the previous French fiscal regime which had a 3% limit on the fiscal deficit. The French regime came from a conversation early in the Miseron government uh where a civil servant called Gabe um was asked by the Treasury Minister to come up with a fiscal regime, and he actually gave an interview about how it happened when he was retiring, and he just said, "Well, you know, the minister came in and said, 'I've got all these people coming into my office wanting to spend money, we need to have a limit on the deficit, I need you to give me a number that's good.'" Sounds like it come from an economist and sounds convincing. And Abe kind of sort of looked up and said, "Well, you know, there's no point in setting it any lower than two because two is what we've got at the moment. And you know, if we were going to say five, then 5% of GDP feels like it's a bit high and won't constrain anything. But we're basically choosing between 3% and 4%." and it goes, "Well, you know, three is a good number. It reminds me of the Holy Trinity. Lots of good things come in threes. Let's say 3%." The whole discussion took an hour. And but once that's in place, anyone who's going to change their mind is going to have to make a very very visible decision. So, similar with the kind of inflation targets. It's pretty well accepted these days that 2% for the UK inflation target is probably the wrong number. You know, everyone who's looked at this technologically goes, "Yeah, we kind of wish that we'd chosen four or even, you know, 4 and a half to 5% as a decent steady state rate of inflation. It would leave us that much more room for stimulus kind of at the, you know, at the lower bound." Why did we choose two? Basically, we're copying New Zealanders. Um, the New Zealanders were the first central bank to have an inflation target. Why did they choose two? Because um the yeah, the Treasury Minister said on offhand television interview that he thought he'd rather see it at 1%, and they reckon that they had a 1% measurement error, so they chose two. And and that that was it. That was again the the story of why it's 2%. But of course, everyone knows that if you move that inflation target from 2% to 4%, you're always going to go down in history as the guy who raised the inflation target. You know, it's very you're going to have to make that. There's no way to kind of secretly make that decision without it attaching to you. And of course, a bunch of people who bought bonds at the old price are going to lose money. So, you're going to be in a very uncomfortable position of having a bunch of extremely rich people with a grudge against you and not really being able to pretend that it's anyone but you that they ought to have the grudge against. So these things just carry on, you know, and here we are closing to 50 from 40 years from GearBay's original decision, and 3% is still being debated in Europe as to whether that's the right number, and everyone kind of knows it isn't. And they're just now bolting on more and more escape clauses and just trying to erode the significance of that without anyone being prepared to bite the bullet and say the whole thing's the whole thing's gone. This the whole stability and growth pact is is gone. It can't be kept saved.

I mean, it's so crazy that it it from a decision-making standpoint, it it was kind of easier to, you know, let let Greece burn and let all these horrible things happen in the EU than for somebody just to be like, "No, I'm the target's wrong based on some actual economics which you could do, I'm changing it." Like, nobody was willing to do that.

I mean, is that the 3% the Maastricht treaty that's an accountability sink, right? Am I am I right that that is?

Yeah. It was like nobody well you know nobody the longest time firstly you know I mean if we go back to the you know the start of where this came in, everyone wanted to be like the Germans. Everyone wanted to borrow German credibility. You know, everyone wanted to be perceived as being basically economically like the Germans. Then you know the world kind of cringed, and suddenly it became everyone started going, "Oh yeah, you know, if your economy is actually that of Greece, it's not such a great idea to have committed to being um a bunch of auto-liberal um anti-inflation cases." You of course Germany was there going, "You know, we you know, nobody was there was no false advertising from the part of the Germans, you were you know, it was part of the whole selling point of the thing that you were going to be in currency union with a bunch of other liberal inflation meters." But no one wanted to have that discussion of how you were going to create a currency which could incorporate these very different economies, what sort of fiscal transfers you would be needing. And then nobody liked the idea of having a big political meeting at which basically Germany and France sat down and decided what the budget was going to be for Greek pensioners. You know, those, you know, because you were always going to end up with either a answer that was economically unsustainable for Greece or you were going to end up with one that was politically unsustainable in Germany. And if you got something that was more politically unsustainable in Germany than what we actually had, the whole euro would have fallen apart. So everyone was just very happy to avoid all of those difficult decisions by saying, "Well, we solved all this 10 years ago. It's 3%. It's the stability and growth packs. It's it's a magic number. You can't break that. You know, what are you crazy? It's a number."

M, so this does seem like a good entry point to talk about like what a good system might actually look like, right, and that's where the the Stafford Beer and the cybernetics come in. So I mean, you know, I I don't know how much detail we necessarily want to go into with exactly what it looks like, but could you sketch like maybe some of the key principles that beer that beer thought made for a good decision-making system?

I mean, high level um, and yeah, it is possible to go down a massive rabbit hole with these things, but high level basically cybernetics is just information theory applied to decision-making in organizational contexts. It's literally a lot of the same mathematics as Claude Shannon and all of the kind of information theory that you see in electrical engineering and telecoms, but applied in contexts where you can't necessarily quantify the number of megabytes per second, but you do know that the feedback connections are either there or they're not. So the big principle of design is always making sure that there is sort of a cybernetic equivalent of accountability. And the cybernetic equivalent of accountability is that there has to be feedback from the consequences of a decision to the decision-making system. And a huge amount of what goes wrong in our economic and political system in my view is that those connections get broken either because they are politically or organizationally inconvenient for someone or because it looks like they are a waste of time and money because they're usually not doing anything, you know, very important. you know, you you you always kind of look at something like the emergency brake in a train and going like, "Well, I hardly ever use that. Let's get rid of it." Um, so when you're designing a system, you know, like the the Europe or like the Must Stability and Control Pact, the decision process isn't finished when you've got what you think is the optimal design. The decision pro the design process is only finished when you've also designed fail-safes and feedback mechanisms and arrangements to say how you are going to change this if you run into a situation

In which it becomes unsustainable or damaging. And it's those fail-safes and feedback mechanisms that are just absent in a lot of mechanisms that have been designed specifically by economists in the neoclassical tradition, because economics in the neocclassical tradition tends to think mostly about perfect information and to think about equilibrium. And so it has this habit of designing systems with really nasty properties in out-of-equilibrium situations or in situations where the information is difficult or costly to communicate. It's so it's this this idea of like levels of the system is is quite key, right? And and how, you know, one each level has different roles and it's it's a kind of the buck stops here. Is that that was Harry Truman, was it? It's kind so I think system five—so there are five systems for people who haven't read the book, maybe we won't be able to explain them all in detail here—but like system five is like the buck stops here system. Five inter is like the ultimate intervention, but you know, system two similarly uh controls system one to a certain extent, and it's it's really interesting, right, how these systems interact, because you mentioned economists and their maximization stuff, and it does just strike me that they don't they just don't acknowledge this at all. Right? There's just there's kind of there's just the maximizing agents. And I guess sometimes in some models there's like a central bank. So there's maybe two systems in some models, but really this is all absent from their frameworks, right?

There's yeah, you know, it's like every most, you know, obviously, it's hard to generalize about economics because there's always someone doing something in somewhere doing interesting work that it's not true of, but there's economic models tend to have this huge information-reducing filter placed on every decision because every agent has got a very, you know, they face an information environment that's entirely made out of prices, and they make decisions that are almost entirely um based on quantities. And so you've got this price and quantity. If you're doing really sophisticated game theory or sort of uh industrial modeling, you might have some kind of quality or product differentiation. Mostly you're dealing with completely undifferentiated goods. Even if you're doing really complicated game theory of the sort that's almost impossible to derive any useful macro conclusions from at all, you might be able to cope with the fact that economic agents have ongoing repeated relationships with each other. Um if you actually kind of think about how pricing, production, and employment decisions are made in the world, you know, the most complicated cutting-edge game theory um model of how you might kind of set the pricing and output of a beer build would be just a trivial problem you'd give to a graduate student uh on the training program with, you know, you know, the actual management of the economy by managers is just so takes so many more things into account, you know, because for the economists they have this high-level philosophy that what they are doing is constrained optimization of basically a single quantity or something can be reduced to a single quantity, and that hugely affects the kinds of things they can think about and the kinds of problems they can solve.

It's interesting because it's it's one of the things that you say is it's like economists are losing a lot of information um but in all the wrong places, right? So it's like in one sense they're making the problem too complex because you've got to optimize. So you got to solve some constrained optimization problem, and those are very limited in what you can actually solve. Um and then so they have to drop everything from it and assume perfect information and so on. Whereas what you're talking about with this cybernetics seems like it's more of a, you know, you said the feedback's like yes/no; it's not like optimization; it's more like, you know, you get feedback or you don't, right, and those are the nos you need to design, and you can imagine some kind of model—I don't know if it's necessary—but like some complexity model or something that, you know, it's just just a series of of, you know, sort of like an electronic circuit, right? It's like and/or, you know, yes/no, and then the feedback continues up the path until there's a loop, and that's not that's not as complex as as optimization, but it is actually more realistic in a sense. Does that does that make sense what I've said?

I think what you described there is that um it vulation by veto is an extremely useful kind of technique for building systems which have to deal with a lot of uncertainty and where you're not sure about how much bandwidth um you need to build in at any point because you just set up the system so that every unit in it, every kind of black box on the map is capable of broadcasting to the whole system: "This current arrangement is unacceptable to me," and then everything kind of changes and just, you know, people grope around for other solutions until all of the boxes are sending the signal "This arrangement is acceptable to me," and that's that's not necessarily going to be an optimum. You know, there's a whole load of textbook maths about how close you might expect it to get to an optimum. And sometimes those arrangements can be pretty good, but it's not going to be an optimum, but it's always going to be a system, an outcome that is acceptable to everyone in the system. You know, the trouble being that you then get strategic behavior. You then get people, you know, claiming that the uh outcome is unacceptable to them when it isn't. And dealing with that is one of the big kind of unsolved problems of politics, which is probably a bit above my pay grade. But it's, you know, it's better than having a system that just goes for constrained optimization of a couple of the parameters and then leaves out huge amounts of things that people care about. Yeah.

It's like and the the it just reminds me again of this like the fact that we don't have that "I'm not satisfied with the system," you know, on a granular level. Um I'm not satisfied with my mobile phone provider. I'm not satisfied with the way the central bank's functioning. We don't have that on a granular granular level. This again is where you get this this just massive backlash. It's just like, well, I reject the whole thing. That's all I've got now. I reject the whole thing. Absolutely, absolutely, and you know that's that's what you see with Brexit. That's what you see with Trump. One one of the things that kind of motivated me to start writing in this area is that there's no point of having a theory of Nigel Farage that has any Farage-specific characteristics to it; your theory of Brexit has to work as a theory of alliance for Deutseland and a theory of Trump and a theory of Modi, um it's because all of these kind of leaders and political movements are so obviously similar that it just it's just not credible to say that we've had, you know, 10 or 12 institution-specific, country-specific things happen that look really so similar and happen at exactly the same time. And so this idea of responsibility really interests me as well. Um because any any one who is at a, you know, a higher level of the system has a responsibility to make a decision, right, and and this is what we're lacking now. This is the unaccountability that we've we've spoken about. Um what I really think what what really interests me here is how again how this relates to economics because I think bringing in the idea of responsibility. It's something that you don't you won't really see that at all in in economics. It's much more there's like legal doctrine surrounding it, right? But it seems to actually although cybernetics sounds kind of techy, I'd say this idea of responsibility is very human, right? And so there's there's you're you're really acknowledging something a bit more human in the way we we design our systems here.

So yeah, you know, it's just you have the decided upon, and there has to be some way to get their views into the rest of the system because the um, you know, I mean my own background is partly economics but mainly business school, finance, and accountancy, and one of the things that really you begin to realize, you know, as you spend a life uh dealing with accounts is the amount of time you spend with a set of accounts trying to include information that's not in the accounting system. Because you know an accounting system, you know, like a set of economic statistics, like kind of anything that we kind of measure in that whole kind of area of our intellectual life is defined by what information it includes and what information it leaves out. Um so a couple of years ago I was involved with—I later left it—a project for bringing carbon accounting into financial accounting. And so basically being bringing climate change risk and climate change factors into financial accounting, which was, you know, it's a great project. There are still people doing it, working on it. I wish them well. But one of the things I thought as I was leaving that group was, isn't it kind of crazy that we've got an existential threat to human life on the planet Earth and the only way that we can deal with it is by bringing it into a financial accounting system. You know, it was like, you know, is it shouldn't we be able to react on that information directly rather than by sort of arranging things so that it affected earnings per share. The reason that that's difficult is firstly, you know, a lot of carbon uh accounting and a lot of environmental factors just don't belong in a set of financial accounts which are there to record financial transactions. But more importantly, we'd set up our whole corporate system of information and incentives so that it only responded to the information that was in the accounts. There was no way of getting people to respond to other important information because we'd set it up so that these things were set up to maximize profits and maximize shareholder value. And if you set up your system so that it's always ignoring certain kinds of information, then you better really really hope that the information that you're ignoring is never going to be something that can damage the system. Yeah.

Yeah. That's definitely definitely not the case in that instance. Right. I I think you know this this leads me to something that I been thinking about which is like it's quite hard to pigeonhole you politically and ideologically. Clearly, you know, you're not a Trump supporter. You're relatively progressive in your values and stuff, but you don't come across as like an overt socialist either. But I do have to say throughout your book and this interview, you know, you've just been telling me time and time again where basically capitalism seems to conflict with the design of a good system. I mean, do do you think that's true? Um, kind of capitalism of the period since 1970, in fact, specifically since September the 13th, 1970, um uh which I'll get is was the date that uh Milton Friedman's essay "The social responsibility of a company is to increase its profits," Yes. Uh was published. So you know, politically I'm always basically a J.K. Galbraith guy, you know, uh he's always been my intellectual hero. So I'm I'm a Keynesian, you know, what's it that J.M.K. said? "When the class war comes you will find me on the side of my own class, the civilized middle class." Uh but you know, it's the whole project of Keynes has always been capitalism, but you're only supposed to blow the bloody doors off. You know, it's always been about organizing a broadly capitalist, broadly free-market structure so that it doesn't generate reactive forces which will blow the whole thing out. Um and Galbraith was very good and very interesting on that. But J.K. Galbraith operated in the US economy of the 1960s, which is an utterly alien intellectual landscape to us today. It. You could probably make a case um in a blog post, probably. I wouldn't like to want to stand this up intellectually, but the US economy of the late 1950s probably has more in common with the Chinese economy of today than with the US economy of today in terms of the way that it was managed and the way that it's interrelated with government, technology, science, and politics rather than being completely based on financial reporting to owners of financial capital. You know, it was it's a it's a very very alien form of capitalism that is still people's mental model of what capitalism should be like. And we've, you know, since the 70s and 80s developed this really weird, really obsessive debt-financed, accounting-based, shareholder value-based version of capitalism in the Anglosphere economies and therefore across most of the world, which is um, you know, in my view just kind of genuinely unserviceable. It's genuinely unserviceable at a kind of planetary level, but also, you know, if you look at life expectancies in the United States of America, dealing with that kind of system seems to be increasingly individually unserviceable for a lot of human beings. Yeah.

I mean, are you would you say that you're someone who very much believes in like the power of ideas? Are you quoted Keynes once, then there's another quote about the the power of ideas, right? Would you say you follow Keynes in that as well? You know, it's the power of ideas, but not necessarily otherwise the power of good ideas. You know, the the last kind of 50 years is really the story of Milton Friedman's version of capitalism winning, and that being very very bad for everyone concerned. Yeah, I mean that's that's fair enough. I think I go back and forth on this because I think somebody who was maybe more on the radical socialist side would would argue that this is this kind of has to happen. This will happen. You know, a Milton Friedman may be the the main procilitizer, right? But like it kind of had to happen because of crises of profitability and the general tendency for the capitalist class to accrue more power and kind of reorganize things in their image. Um and so that was kind of an inevitability. Neoliberal capitalism, maybe not the exact form, but something like it was an inevitability um for the for the US. I mean I mean do you do you disagree with that? What we I would describe it as as the kind of the invisible class war uh as what the kind of 1960s paternalist big-company managed capitalism of J.K. Galbraith's era was a system that people were beginning to identify as managerialism uh because you had what was you know basically a free-market economy, but the means of production were controlled basically by managers uh by the tech what called the technos structure uh but basically by specialists administrators who formed a class that went through business, government and the army and just arranged things um subject to fairly minimal oversight, right, by investors, elected representatives, and kind of the sort of nominal elite. From the 1970s onward, you had a massive class war between the managers and the capital owners. And it was one in which the capital owners absolutely won, and they absolutely won by decimating the ranks of middle managers um massively culturally stigmatizing them. No one wants to be called a middle manager these days. Um and you know, at the end of this we have this weird neoliberal non-managerial capitalism where everyone is meant to be managing according to financial markets. Um a lot of which was driven by people presuming that financial financial markets were the ultimate information-processing system that you could just throw any question and you would get an answer coming from the market about anything that you wanted to do. So is that like does that relate to the the efficient markets hypothesis, would you say is that like the the epitome of that?

Exactly; it's just the this kind of belief that you can judge any policy from a government by looking at bond yields, and or you can, you know, horrifically in the case of Boeing, you can make all of your judgments about the design of a new airliner by looking at the share price. Yeah. Um and you know, the trouble is there again it kind of comes back to cybernetics and information theory. The fact that a stall sensor is badly uh installed and needs a duplicate and will cause uh crashes just fundamentally isn't a piece of information about the stock market. It just fundamentally isn't a question that you can expect the equity market to answer for you. Um similarly, you know, the future of the European Union, you can't it's just such a complicated thing that you can't expect a single number which also has to do the job of clearing the uh savings and investments uh of the bond market to give you all of that information. I mean, you you'd be talking about moving away from that, you'd be talking about some pretty serious institutional changes, right? You'd be talking about like completely different ways of like financing governments and and maybe big restrictions on the amount of trading that you can do, or is that the type of thing you'd have in mind?

Honestly, I think that you could achieve a lot by getting rid of—I'll say getting rid of debt—and then I'll immediately qualify it by saying that there is some kind of situations in which companies and individuals do need to take out debt, etc., etc.—but hugely reducing, by which I mean rolling back probably the last 40 or 50 years worth of growth in the use of debt, because most companies should be debt-free. Most organizations should basically be financing themselves from retained profits rather than borrowing loads of money. It should be hugely regulated the extent to which you're able to take over a company using borrowed money. Um weirdly, actually, since you've written the book, I've started appearing on a lot of sports podcasts. Uh because some of the people who are most interested in understanding what happens when you take over an entity with borrowed money and change its incentive structure are football fans, uh particularly Manchester United fans. But um you know, it's that whole practice of, you know, the modern private equity industry as it is just basically, you know, I've I have friends who work in that industry, and I've said this to their faces: I think it's all bad. I think there's nothing that can be saved from it. Um in general, you probably don't need to do that much to the structure of government financing in the G7 industrialized nations, for emerging market finance. I mean, I don't want to sound like some kind of uh mad Sinophile here, right? But you know, we do have a very interesting model of the way in which China interacts with developing countries versus the way in which the western Anglosphere capital markets interact with developing companies. And although I'm certainly not going to co-sign, you know, or any reputation of the People's Republic of China and how it eventually ends up dealing with its African client states, I think everyone in those African governments would probably agree that they've had a better time dealing with uh China for the last 20 years than the time that they've had dealing with bond markets in the 20 years preceding doing that.

And what what is this model? What's what's the key difference with this model? It's basically non-market fiat, you know, it's um you've got, you know, and there has been crises. Um I I was born in Zambia, so I follow the Zambian news every now and then, and so kind of aware of this switch in the relationship there. Companies which run up big debts with PRC for development projects do get into trouble, and they do have debt crises with respect to their PRC debts in the same way that they used to with their bond market debts and with their IMF debts. The difference is again you've got a richer channel of communication because it is a state-to-state, head of government to—well, not exactly head of government to head of government realistically—but head of government to important guy in the uh finance ministry of the Republic of China. You can have richer communication and make longer-term binding deals than in a bond market system where really all you can do is set a price in a quantity, and you're left, you know, when things get into trouble, you're left trying to negotiate with a very aggressive ad hoc committee of uh bond holders generally run by bankers with much less skin in the and much less commitment to any kind of longer-term relationship. So it's I kind of feel like it's significant changes but not necessarily changes that would leave the whole system unrecognizable that are needed. But it's conscious management. It's like conscious understanding that when we're doing these things, particularly in the world of finance, we're talking about long-term projects. And so we've got to build institutions that are capable of making long-term decisions. And this is a pretty radical change. It's a very interesting idea. It's also completely in line with what you're talking about, right? Because it's um, you know, China is accountable, right? In a way that the market's not—the bond market's not really accountable—and that's something which just makes it extremely difficult for people to exert political will over over the decisions of their governments. Yeah, absolutely. And I think that's, you know, when obviously if you say a phrase out loud like "China is accountable," you're immediately

Vulnerable to the objection. It's not be silly, but it's it has that feedback link in there. You can communicate with it in a way that you can't communicate with the bond market, and a way that you can't communicate with the European Central Bank, because you know all of these institutions were designed in many ways not to be communicated with because they were designed because people felt like they'd had about enough of listening to input and feedback from the people who were bearing the consequences of their decisions.

Yeah, there's a kind of shush we're calculating um to it, isn't there? I think it's like Yeah, it's it's really interesting. It does bring it does it does you know you talked about having a theory of everything. It does seem to bring a lot of things together.

Well, it you know like any theory of everything, it's a theory of nothing because it doesn't work until you start building actual institutions. But I've you know the one thing I kind of want to say in the book is that all decision-making institutions are information processing institutions, and they have to be designed that way. You have to make sure that the information balance sheet is going to balance uh in terms of the complexity of the environment with the capacity of the thing that's managing it. You know, otherwise you get, you know, you get a Donald Trump situation. If you have Donald Trump in power, then you have to reduce everything to a complexity that Donald Trump can deal with. And then you have to work out how you're going to change his decisions into something that isn't going to blow the system apart. Because if you can't do that, then the system will become unregulated. And that's just uh unfortunately um for those who are kind of still playing along at home, that's theorem 9 from Claude Shannon's theory of information. Theorem nine.

Okay. Yeah. If anybody's got the the theory of information open then.

Yeah. Yeah. Can cross-reference that. Um don't ask me for a page number or else no let actually I really want to talk to you about Stafford Beer the man cuz I I think there's a lot on Well, there's a whole chapter on him, right? He was an interesting guy and you seem quite fascinated by him. So could you kind of talk us through like how he came to this and and what his main ideas were?

Yeah. Uh well like Stafford Beer um if you it's very interesting Google image search will tell you a lot of Stafford Beer's biography because half the pictures of him are of a kind of big guy in a double-breasted suit with a cigar from his period as a management consultant. And then a lot of the rest of the pictures of him are much longer beard, flowing robes. You know he went from management guru to actual guru basically over the space of 15 years between 1965 and 1980 he was just this slightly you know clearly incredibly intelligent intuitive mathematician who was one of the few people who took Norbert Wiener's book cybernetics which was you know the invention of information theory based on uh the war work on designing uh automated bomb sites and things like that. He was one of the few people who really understood that and he started applying it to management problems starting at a variety of British steel companies in the 1950s. Um again a whole different world. British steel companies were at that point the envy of the world. They were extremely well-managed. He left um the well one of the precursor companies to British Steel. um worked for uh in publishing for a while, worked as a set up his own management consultancy, but he was just generally kind of always trying to help companies organize themselves so that their internal communications would just respect the basic principles of information design and making sure that every stage age information was being represented which respected the compromise you need to make between the complexity of the world and the capacity of the management team to deal with it was kind of the sorts of things that he did he kept on writing books kind of detailing how you would put this kind of framework into practice and he was just kind of getting to the height of his powers. And it would have been really interesting to see how he dealt with the 1970s and the widespread availability of electronic computing in uh corporate settings. But then Chile happened and what happened in Chile is that um he was called in or called up by some Chilean engineers who were had been recruited into the economics ministry of the Allende government and asked to come over and help them build you know a new economic form of organization. And this was, you know, just the most exciting kind of project of his life. He went over to do this and they were starting to create a network for running the national industries of Chile. Is like a lot of people kind of have preferred to pretend that they were trying to make Soviet planning work with computers uh which absolutely wasn't uh what they were trying to do. They were trying to introduce a communication network that would delegate authority down to the lowest level where the problem could be solved. Then provide a kind of feedback network so that issues as they arose on the factory floor or at the industry level could be escalated to the central government to be solved as they arise. And you know it was really it looked like a fascinating experiment. Obviously then they had the Pinochet coup in 1973 so it never really even got a chance to fail. Uh but that was that was the story of Stafford Beer and that completely derailed his career in management science. From then on in, he spent three or four years trying to ensure the physical safety of the Chileians who had worked with him. And then he started going into philosophy and writing really kind of strange and incredibly interesting textbooks that kind of they're like it's like half kind of yoga and meditation. And then every now and then you'll get a chapter that is nothing but equations because he feels like it's necessary to do some stochastic control theory in order to understand the next bit of uh so yeah so it's um it's a real rabbit hole. Um there's uh he his books are quite expensive so only get them if you see them in a charity shop. Uh but they're always there's always some really interesting stuff in there.

I mean with with Chile I think I was wondering if you could speak to it a bit more because like cyber syn it was called cyber syn right that's like the one a lot of people talk about a lot of people ask me about like what I think of cyber syn and I suppose in one sense there's not too much to say because it never really got off the ground and it was you know you know literally sort of bombed or whatever and that and but but I mean do you think did does it seem like it was a a a good attempt at an implementation of the the principles that we've been talking about.

The big thing about it is that it was an attempt to impose an economy-wide system of management by exceptions. Yeah. So it really was absolutely meant to implement the principle of delegating everything to the lowest possible level because of course the lowest possible level of decision-making is the one that's in closest contact with the reality on the ground. Yeah. So it was an idea that would actually hopefully dispense with systematically connected government statistics and replace financial accounting because you would no longer be attempting to collect all of the information about production and centralize it in one place. You know the idea was that all the problems should be dealt with at the level where they happened. That meant ironically that there was a control room. Um, and the control room is the one that you can see pictures of on the internet. It looks like it looks like Star Wars. There was apparently considerable design influence from Star Trek, but unlike Star Trek, the chairs have got stray ins.

Okay. Um, uh, but the control room was in many ways misnamed. It was this meant to be the central problem-solving room and it would be a room where you brought various kinds of technical experts to deal with problems that had been elevated all the way up to the center because they couldn't be solved at the shop floor or the regional levels and required um kind of action from central governments. We I think that that's quite an attractive model for all sorts of organizations. You know, there are plenty of companies that work like that. I was lucky enough to work at a couple of investment banks that pretty much worked exactly like that uh in that we had a partners committee that met every week purely to discuss problems that had been elevated because they couldn't be solved anywhere lower down in the bank. And you know, could you run a country like that? I think yes. Would you run a country in Chile's specific situation, which is to say one of the big targets of the USA during the Cold War, um, with a mining industry with a very substantial reactionary uh, kind of movement and with an computer network that basically consisted of 400 telex machines, it didn't feel like it was getting a fair shot, but uh, I think that, you know, Idi Medina's book on Cyber Syn is fantastic. Um I think she takes a much more cynical view of it uh than I did uh after reading her book. I think it really could have got somewhere.

Yeah, it's that just like that progressive escalation, right? This is what what we're talking about. It's like a problem on the shop floor gets escalated to the managers and then most most of the time the managers can solve it, right? And then but then if they can't they escalate it to the next level. So you've got like you know only one 1,000th of all the problems that come in the shop floor are going to end up at the highest level right and so you just end up like yeah with only the the most powerful decision-makers um are given the most unsolvable the most important problems and it in such a sensible idea um I mean you said you've seen it in in in business have is there any other case like countrywide or a bigger level than business where you've seen something like it um not really so much ex I suppose you know my own experience is all in banking and weirdly the central banks of the world have a club that works quite like this for their organization you have the bank for international settlements which the BIS um the international settlements in uh in question are German First World War reparations. Um, you might wonder why is a bank that was meant to facilitate German reparations for the First World War still around in 2025. And the answer is that it just proved to be very convenient. Um, the First World War reparations of Germany involved substantial transfers of gold and loans between central banks. And so they had this little club where they always met in Basel. And over time it just became extremely useful for them to keep on meeting there. And big problems like the uh 1980s um Latin American debt crisis, like the rise of uh Japanese banks financed by domestic real estate gains um taking an oversized international role in the 1990s like the global financial crisis of 2008. All of these things ended up being escalated to the BIS in Basel where you have the Mark Carney's and Andrew Bailey's of the world just really sitting down in a room and thrashing the problems out. It's a very interesting strange organization but and it's very dependent on quite close personal ties and ideological sympatico between central bankers. But there are a lot of things particularly in the global payment system that have just got worked out there. And in many ways it is sort of a model that everyone in the world who runs a central bank is meant to take care of their own garden. But when there's things which they find can't be dealt with one by one, they all get together in a little club to sort it out in Basel.

Yeah. Okay. Well, that that that that seems like it's a sensible part of central bank policy, right? We've been talking about some unsensible parts of it, but it's definitely seems to make sense. I've got to ask you about central planning. Um because um about half of my audience for some reason seem to like central planning despite its obvious failure. How does it relate to your to your framework? This framework in particular, would you say?

Well, you know, the thing is there's two problems with central planning, which is that it's planning and that it's central. Um, and you know, the problem with planning is always getting the information to where it can be processed, turned into a decision, and sort of sent back out into the world quick enough to be able to respond to the system to to the changes. And that is a technological problem, but it's also an accounting problem. You know, whenever you're trying to deal with central planning, you're always putting that information-reducing filter onto things because all of the reality of the world has to be compressed down to a model of the world that the central planning authority can deal with, you know, and that is to my mind it has almost all of the problems of financial capitalism. um plus a few more because once more you are just washing everything down into a reporting system which involves throwing away huge amounts of information. Uh but then you are usually processing it in a much kind of smaller information processing system even than the financial markets which I've already said I don't think are adequate to the task. What you need is something that's more like a decentralized planning system, you know, where you just have uh what Stafford Beer called the resource bargain where every economic entity gets given effectively a budget and told you can do what you like as long as you do not make excessive demands on the rest of the system. You know, and some of those demands might be in terms of resources that uh you require, but also you might say you have a budget in terms of the amount of input from central government that uh you are allowed to take up, the amount of political unrest and local opposition you are allowed to create. And just setting that kind of bargain of the resources that entirely decentralized um kind of planning units can make and then dealing with exceptions to that and then having a central planning system being more of a central dispute reconciliation system whenever those bargains need to be renegotiated or whenever kind of resources need to be transferred from one unit to another. So yeah, I kind of I can see something much more democratic than a market system working, but I don't think it would look anything like any of the examples of central planning that we have. And I specifically, you know, if we're going to kind of fire some shots here, I don't think it would look like the kind of linear programming supercomputer that Paul Cockshott thinks is viable, you know, or any artificial intelligence powered version of saying.

Yeah. I mean is is this is this abstraction from reality and you are going to simplify some elements of it and the the the problem is is very difficult to solve as well. There's some people who think that maybe that that can be done with modern computing techniques, but I don't think you're getting away from the fact that you're fundamentally turning something into a model uh with central planning. And any model will abstract from reality. And then you know the bigger the model the and the bigger the plan the more authority the planner has imposing this model then the more absurd some of the outcomes are going to be when you realize that you know you haven't matched onto the nitty-gritty of of the real economy of real real everyday life.

Yeah. Absolutely. And also you if you're building the central planning system um as as I keep on saying you've also got to build a central feedback system. You've also got to build some kind of system for carrying information right back up from the population in exactly as much resolution as you're doing the planning in order to make sure that you're not having uh you know one of the one of the reasons that the market system has worked so well is that there's always the ability of a customer to transmit one bit of information by refusing to buy something. Yeah. you know, and uh you know, you can in a kind of old-fashioned sort of Soviet system, you don't even have that single bit of information that's going up there.

I think there's there's sometimes a bit of a conflation between planning and central planning, isn't there? Cuz what you're describing that you'd like to see is quite clearly a system of planning. There is a plan and it's kind of central, right? But the point is that it doesn't include everything. There's decentralized elements to it, substantial decentralized, and there's feedback. Whereas I think with some of the central especially 20th-century central planning, it's like this is what we're going to do to you and and that's it, right? So so there's there's nothing that we're talking about that we need here. And and almost all of those 20th-century planning systems were designed at a time where it was clearly technologically impossible to have any of that feedback in there, you know. So they were always they they were you know we can see with the benefit of hindsight that this was creating a system that was bound to be fragile because it was incapable of responding to its environment.

Yeah. Yeah. It would it was Yeah. It was created at a time when it was Yeah. even worse than it than it than it would have been now. Right. Yeah. Um, let me ask you. So, I'm I'm at the stage of the interview where I'm asking about all the buzz uh topics, all the buzzwords, but you you have been contributing to Crooked Timber, which is the oldest blog in the world as far as I'm concerned. Um, it's been going for ages. So, I know that you know Henry Farrell and Cosma Shalizi right now. They've been doing a lot of interesting stuff on artificial intelligence and they basically their argument is that artificial intell uh well actually no let me rephrase that their argument is that markets are already a kind of artificial intelligence right they're already a sort of social feedback loop and artificial intelligence it's it's a misnomer we all know that but what it does is it takes what we know statistically kind of averages it and and spits out like some some kind of normalized version of what we know Right. Um, but they say that markets kind of do this uh do this as well. Now I I don't know. This feels like it links to your book. I really This is a very vague question. I don't know how. Does it link to your book?

No, I think I I do know Henry Ome Henry uh actually gave me some very uh useful comments on the the book. Um and it's a very kind of information theory cybernetics way of thinking about these things. Um and artificial intelligence is a horrible word because we have so many connotations to the word intelligence. But it is a decision-making system, you know. So you you know you can have a computer and it the decisions it makes they're real decisions. You know they can turn you down for car insurance. um like a artificial intelligence kind of an AI program you know markets will take in information of various sorts and they will output decisions of various sorts. So if you think about it purely as a black box they are doing the same kind of thing you know and you know you can you know you can tune the computer to maybe have different priorities or spit out different kinds of decisions. On the other hand the market unlike the computer is capable of searching out different forms of information. So there are different kinds of decision-making systems but yeah you know these are these are decision-making systems you know and a a market will take in information and put out uh and and have output and is capable of uh accepting feedback in exactly the same way as a computer system can can be. So yeah and I'm totally on board with that.

Okay. Okay. Yeah. Interesting. Um, yeah. I I guess it it really reframed how I how I think about AI. I mean, is it Could you see a role for that cuz you called it you your system 4 is intelligence, right? You call it intelligence. Could you see a role for the for this, you know, mo what we call AI in in these kinds of systems? Do you think they would they would massively improve it?

It's the whole the whole history of organization is that the world gets more complicated at a multiplicative rate basically exponentially because as you add more things the number of connections between things grows multiplicatively. Your ability to manage that complexity grows basically linearly. You you add more management, you add more resources. This means that over time the gap between the complexity of the world and the capacity to manage it reaches crisis point something becomes unmanageable and we have to reorganize you

Know, and if you take something like Alfred Chandler's, uh, kind of books on the history of management, this is an absolute constant. This is what companies do: things get too complicated for them, they reorganize, and that reorganization has always involved a form of devolution. So the central management gives up on trying to administer at a particular level of detail and delegates things down to, uh, subsidiaries or multi-divisional companies and so on.

Within that linear growth of the capability of management, we have occasional step jumps up with technological revolutions. So, you know, like things like writing or double-entry bookkeeping, you know, I have this kind of fantastic kind of example that me and Henry both tried to use in talks, which is that in its day even the humble filing cabinet was a big jump in information technology. Uh, it was; it made; made the railroads possible in many ways.

What we've got here with your modern generative AI systems is something that really looks like a technological leap that would buy us a bit of time to work out how we are going to reorganize our economy and society because we really need to. The issue is how do we use that?

Um, there is this Stafford Beer joke that I try to get into every presentation I do, which refers to his experience as a consultant at the end of the 1960s when corporations were just installing IBM mainframes. You know, this was something they could have used to completely revolutionize the way that they tracked information from marketing to production to kind of customer relationships and feedback. There was a lot they could have done, but in fact, it would have been politically and organizationally difficult; it would have put too much pressure on their existing management relationships; it would have disrupted existing hierarchies. So what they generally did was install a pay, uh, mainframe and used it to automate their payroll or used it to automate the production of their quarterly accounts. Our Stafford Beer said at the time it was like being given the chance to hire Galileo, Einstein, and Mozart and then put them in a room memorizing phone numbers, so that you could memorizing the phone book so you could look things up more quickly.

M, we've got with generative AI something that's clearly capable of doing very surprising things in terms of summarizing huge amounts of information and giving a different kind of access to huge amounts of training data, and it looks like we're going to use it to send more emails. You know, it feels; it feels like a terrible waste of capacity.

So, you know, I think the answer to your question then, Go, is yes. So, I think it could; um, I think the way that we currently seem to be gravitating towards using it, uh, could be really, you know, a massively wasted opportunity, and; and I mean you could go further than that, right? Cuz it seems like in some cases AI might be used as an accountability sink, right? If you're just like; if it's being used for certain decisions and you say, "Well, you know, this is just what; just like this is what the market said. This is what the AI said," and then…

Yeah.

Yeah.

Yeah.

No, some policy makers are very alert to that. Um, the last AI act from the, uh, European Union basically put in place what they called a right of explanability. So say that if any decision affecting your life is made by a machine system, then you have a right to ask, you know, to have that decision explained to you and effectively kind of regulating the way, you know, the building of such systems so that you can't build one in such a way that it can't be; can't give human-relevant explanations. Um, which, uh, I thought was a really good kind of step against something that would really obviously be a kind of tactic open to abuse by corporations.

Um, I did manage to irritate my friend at the European Commission though because when the explanability act came out, I pointed out that it had come from the European Commission and they asked, you know, when you make a decision that affects, uh, people, what right of explanability do they have from you? And the answer is: something along the lines of "Stop trying to be clever."

Um, yeah.

No, that's it.

Yeah, the EU.

Yeah.

Designing explanability criteria for AI systems but not for themselves. That; that seems like; that seems like a good; a good place to round the conversation off because we've kind of come full circle.

Yeah.

Yeah.

You know, um, I think, um, no, it's; it's; it's really interesting. I actually, there is one final example that I; I wrote down but I haven't said anything to you, but I think a; a good example of this sort of rejection of the lack of accountability and this, you know, "I don't like what's going on," is, of course, um, Luigi Manion and; and Brian Thompson and; and that happened since your book came out. I'm not trying to get you into trouble here, but I just; I what do you make of that, cuz that was; that occurred to me as an example.

Yeah. I mean, I was; I was at a conference in, uh, where was it? Baltimore, uh, when that news hit and, uh, I kind of had about like five or six different text messages coming on because I'd; I'd written something, well, you know, when you; when you shut down all other forms of feedback, you know, the; you; a really restricted communication channel, the only message that that channel can carry is a stream, and I've actually written that in the context of Trump's election, and six or seven people go, you know, sort of started throwing that quote back at me and linking to this headline, and I was thinking, oh, what if there's an opportunity to get myself in trouble, you know, there was, you know, not to; not to, you know, not to condone any activity, and violence is never a solution to anything, but this kind of thing did not used to happen. Um, chief executives did not used to need the degree of security that they have.

Yeah. Part of the problem here is that there is no way to communicate with companies in general. Insurance companies have a particularly bad reputation and track record for being difficult to communicate with. American health insurance companies more than anyone else. Um, you know, it's, you know, group hoping that things like this don't become commonplace because I do kind of like living in a society where you can walk down the street and there aren't armed guards everywhere. Yeah. But yeah, you know, it's, uh, it's a sign of the kind of world that we're living in that lots of places have security, um, where what they actually need is communication. Yeah, people; people need to be listened to. You know, it's; that's one of the obvious explanations for the incredible popularity of, uh, chatbots and AI. You know, not feeling listened to is one of the big issues of our age. And you know, as; as anyone who's brought up a toddler or, um, been in a bad relationship knows, feeling like you're not being listened to is one of the most psychologically intolerable things on earth.

Yeah, that's very; very true. And I mean the other thing about it is that it doesn't; despite the act, probably doesn't make a difference to, you know, I don't know; like it doesn't, you know, the CEO; you talk about this; you say, not when you shoot the CEO, but even if you manage to get the CEO on the phone, they probably wouldn't be able to do anything either. So it's like; it's a; it's an act that doesn't actually result in any kind of improved outcome, but it's just predictable; right? It's predictable. You; you; that is actually the other thing that is not necessarily emphasized by the, uh, people who kind of here worship Luigi Man, which is, you know, what major changes in policy with respect even to that health company actually affected. So, you know, it's; I'm saying there are no other forms of effective communication, but going out and assassinating the CEO is also not an effective form of communication because the organization is the organization. And it's the system, not the human being.

Yeah.

Yeah.

And you need absolutely radical overhauls of the system to make it work for people so they feel listened to and they; and they are listened to as well, and there are changes. Um, so yeah, I mean, thanks so much for coming. So the unaccountability machine is great. It's such a great book, and I think, again, I said this at the start, cybernetics, um, the ideas of Stafford Beer, all of this stuff; it's; it's; it's time has come. I feel it's like we really need to be talking about this. I; I kind of feel hopeful that more people are using the phrase "the purpose of a system is what it does," and like if one in a hundred of those starts kind of looking onto the internet to say that's an interesting phrase. I wonder who first said it. Uh, they will end up going down the Stafford Beer rabbit hole.

Yeah.

Yeah, that's good.

Yeah.

The purpose of a system is what it does. That should be household. Everyone should be saying that. Um, cool. All right. Well, thanks. Thanks then. Um, I'll see you later. I'll see you again. Thank you. Have a great rest of the day. [Music]